Steric and electronic effects in N-heterocyclic carbene gold(III) complexes: An experimental and computational study

Citation: Rosero-Mafla, M.A.; Zapata-Rivera, J.; Gimeno, M.C.; Visbal, R. Steric and Electronic Effects in N-Heterocyclic Carbene Gold(III) Complexes: An Experimental and Computational Study. Molecules 2022, 27, 8289. https://doi.org/10.3390/ molecules27238289 Academic Editor: Zhongning Chen Received: 7 November 2022 Accepted: 22 November 2022 Published: 28 November 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). molecules Article Steric and Electronic Effects in N-Heterocyclic Carbene Gold(III) Complexes: An Experimental and Computational Study Miguel A. Rosero-Mafla 1, Jhon Zapata-Rivera 2, M. Concepción Gimeno 3,* and Renso Visbal 1,4,* 1Departamento de Química, Facultad de Ciencias Naturales y Exactas, Universidad del Valle, A.A. 25360, Cali 760042, Colombia 2Departamento de Química, Facultad de Ciencias, Universidad de los Andes, Cra 1 No 18A—12, Bogotá111711, Colombia 3Departamento de Química Inorgánica, Instituto de Síntesis Química y Catálisis Homogénea (ISQCH), CSIC-Universidad de Zaragoza, 50009 Zaragoza, Spain 4Centro de Excelencia en Nuevos Materiales (CENM), Universidad del Valle, A.A. 25360, Cali 760031, Colombia *Correspondence: [email protected] (M.C.G.); r[email protected] (R.V.) Abstract: A series of neutral acridine-based gold(III)-NHC complexes containing the pentafluorophenyl (–C 6 F 5 ) group were synthesized. All of the complexes were fully characterized by analytical techniques. The square planar geometry around the gold center was confirmed by X-ray diffraction analysis for complexes 1 (Trichloro [1-methyl-3-(9-acridine)imidazol-2-ylidene]gold(III)) and 2 (Chloro-bis(pentafluorophenyl)[1-methyl-3-(9-acridine)imidazol-2-ylidene]gold(III)). In both cases, the acridine rings play a key role in the crystal packing of the solid structures by mean of π – π stacking interactions, with centroid–centroid and interplanar distances being similar to those found in other previously reported acridine-based Au(I)-NHC complexes. A different reactivity when using a bulkier N-heterocyclic carbene ligand such as 1,3-bis-(2,6-diisopropylphenyl)-2-imidazolidinylidene (SIPr) was observed. While the use of the acridine-based NHC ligand led to the expected organometallic gold(III) species, the steric hindrance of the bulky SIPr ligand led to the formation of the corresponding imidazolinium cation stabilized by the tetrakis(pentafluorophenyl)aurate(III) [Au(C 6 F 5 ) 4 ] − anion. Computational experiments were carried out in order to figure out the ground state electronic structure and the binding formation energy of the complexes and, therefore, to explain the observed reactivity. Keywords: Au(III); NHC; acridine derivatives; TD-DFT; pentafluorophenyl group 1. Introduction Since the first isolation one decade ago, N-heterocyclic carbenes (NHCs) have been treated as novel ligands in multiple investigations [ 1 ]. However, nowadays NHCs are among the most useful ligands, being used in a great variety of applications. There are plenty of reviews covering different areas in which NHCs play a key role [ 2 – 5 ]. Medicinal chemistry is still one of the principal research areas that takes advantage of the versatility of such ligands to explore new alternatives for the treatment of cancer [ 6 – 8 ]. In 2012, our research group reported the synthesis and characterization of several silver(I)- and gold(I)- NHC complexes that displayed blue–green emissions thanks to the presence of an acridinebased chromophore [ 9 ]. Later, in an extension of that work, various gold(I) complexes incorporating bioactive molecules and acridine-based NHC ligands were reported to be good antitumoral agents against two different cancer cell lines, A545 and MiaPaca2 (Figure 1) [10]. Molecules 2022,27, 8289. https://doi.org/10.3390/molecules27238289 https://www.mdpi.com/journal/molecules Molecules 2022,27, 8289 2 of 16 Molecules 2022, 27, x FOR PEER REVIEW 2 of 16 Figure 1. Acridine-based NHC complexes previously studied. One of the major limitations observed for acridine-based M-NHC complexes is their poor solubility in moderate polar solvents. In general, the incorporation of bioactive molecules such as 2-mercaptopyridine or thio-β-D-glucose tetraacetate improves their solubility and therefore their biodistribution in the cell interior.[10,11] Fluorine-containing groups are known for enhancing polarity when introduced into non-polar molecules. Trifluorophenyl (-CF3) and pentafluorophenyl (-C6F5) groups are particularly well-known in this regard [12–15]. This pentafluorophenyl fragment has also been used to prepare many group 11 transition-metal complexes [16–21]. Its great sigma-donating properties and aromaticity help to stabilize metal ions, which in principle display some instability. For example, gold(III) is typically reactive/unstable under physiological conditions induced by intracellular redox reactions [22]. Taking this into account, the incorporation of the pentafluorophenyl group together with the presence of an appropriate NHC ligand could improve the stability of gold(III) species in physiological conditions for biological purposes. In fact, there are several works reporting on the synthesis and characterization of stable C6F5-containing gold(III)–carbene complexes (Figure 2) [23–30]. N N R N MCl N N R N Au S OOAc OAcAcO AcO N N R N M N N R N N N Me N Au S N N N N N N M = Ag, Au BF 4 Figure 1. Acridine-based NHC complexes previously studied. One of the major limitations observed for acridine-based M-NHC complexes is their poor solubility in moderate polar solvents. In general, the incorporation of bioactive molecules such as 2-mercaptopyridine or thioβ -D-glucose tetraacetate improves their solubility and therefore their biodistribution in the cell interior [ 10 , 11 ]. Fluorine-containing groups are known for enhancing polarity when introduced into non-polar molecules. Trifluorophenyl (-CF 3 ) and pentafluorophenyl (-C 6 F 5 ) groups are particularly well-known in this regard [ 12 – 15 ]. This pentafluorophenyl fragment has also been used to prepare many group 11 transition-metal complexes [ 16 – 21 ]. Its great sigma-donating properties and aromaticity help to stabilize metal ions, which in principle display some instability. For example, gold(III) is typically reactive/unstable under physiological conditions induced by intracellular redox reactions [ 22 ]. Taking this into account, the incorporation of the pentafluorophenyl group together with the presence of an appropriate NHC ligand could improve the stability of gold(III) species in physiological conditions for biological purposes. In fact, there are several works reporting on the synthesis and characterization of stable C6F5-containing gold(III)–carbene complexes (Figure 2) [23–30]. Most of these gold(III) complexes have been proven to be stable enough to be used as precursors in different catalytic reactions and also as anticancer agents against several cancer cell lines [ 22 ]. The aim of this work was to synthesize and characterize acridine-base gold(III)-NHC complexes and to explore the influence on the structural and electronic properties of the pentafluorophenyl group. Molecules 2022,27, 8289 3 of 16 Molecules 2022, 27, x FOR PEER REVIEW 3 of 16 Figure 2. Several gold(III)–carbene complexes containing the pentafluorophenyl group. Most of these gold(III) complexes have been proven to be stable enough to be used as precursors in different catalytic reactions and also as anticancer agents against several cancer cell lines [22]. The aim of this work was to synthesize and characterize acridinebase gold(III)-NHC complexes and to explore the influence on the structural and electronic properties of the pentafluorophenyl group. 2. Results and Discussion 2.1. Synthesis of the Au(III) Complexes with NHC Ligands Derived from Acridine The 1-methyl-3-(9-chloroacridine)imidazolium chloride ([IMeAcr-H]Cl) salt used as a precursor of the N-heterocyclic carbene ligand was obtained according to the procedure reported by Gimeno and co-workers[9] and then reacted with an equivalent amount of the corresponding gold(III) derivative, ([AuCl3(tht)], [Au(μ-Cl)(C6F5)2]2, or [Au(C6F5)3(tht)]), to generate the gold(III)-NHC complexes 1–3, respectively (Scheme 1). Scheme 1. Synthesis of gold(III) complexes 1–3 containing the acridine-based NHC ligand. tht = tetrahydrothiophene. N N CH 3 Cl N [AuCl 3 (tht)] CH 2 Cl 2, K 2 CO 3 N N CH 3 N Au Cl Cl Cl CH 2 Cl 2, K 2 CO 3 1 / 2 [AuCl(C 6 F 5 ) 2 ] 2 N N CH 3 N Au C 6 F 5 Cl C 6 F 5 N N CH 3 N Au C 6 F 5 C 6 F 5 C 6 F 5 CH 2 Cl 2, K 2 CO 3 [Au(C 6 F 5 ) 3 (tht)] 1 3 2 Au Cl C 6 F 5 C 6 F 5 Cl AuC 6 F 5 [AuCl(C 6 F 5 ) 2 ] 2 = C 6 F 5 H Figure 2. Several gold(III)–carbene complexes containing the pentafluorophenyl group. 2. Results and Discussion 2.1. Synthesis of the Au(III) Complexes with NHC Ligands Derived from Acridine The 1-methyl-3-(9-chloroacridine)imidazolium chloride ([IMeAcr-H]Cl) salt used as a precursor of the N-heterocyclic carbene ligand was obtained according to the procedure reported by Gimeno and co-workers [ 9 ] and then reacted with an equivalent amount of the corresponding gold(III) derivative, ([AuCl 3 (tht)], [Au( µ -Cl)(C 6 F 5 ) 2 ] 2 , or [Au(C 6 F 5 ) 3 (tht)]), to generate the gold(III)-NHC complexes 1–3, respectively (Scheme 1). Molecules 2022, 27, x FOR PEER REVIEW 3 of 16 Figure 2. Several gold(III)–carbene complexes containing the pentafluorophenyl group. Most of these gold(III) complexes have been proven to be stable enough to be used as precursors in different catalytic reactions and also as anticancer agents against several cancer cell lines [22]. The aim of this work was to synthesize and characterize acridinebase gold(III)-NHC complexes and to explore the influence on the structural and electronic properties of the pentafluorophenyl group. 2. Results and Discussion 2.1. Synthesis of the Au(III) Complexes with NHC Ligands Derived from Acridine The 1-methyl-3-(9-chloroacridine)imidazolium chloride ([IMeAcr-H]Cl) salt used as a precursor of the N-heterocyclic carbene ligand was obtained according to the procedure reported by Gimeno and co-workers[9] and then reacted with an equivalent amount of the corresponding gold(III) derivative, ([AuCl3(tht)], [Au(μ-Cl)(C6F5)2]2, or [Au(C6F5)3(tht)]), to generate the gold(III)-NHC complexes 1–3, respectively (Scheme 1). Scheme 1. Synthesis of gold(III) complexes 1–3 containing the acridine-based NHC ligand. tht = tetrahydrothiophene. N N CH 3 Cl N [AuCl 3 (tht)] CH 2 Cl 2, K 2 CO 3 N N CH 3 N Au Cl Cl Cl CH 2 Cl 2, K 2 CO 3 1 / 2 [AuCl(C 6 F 5 ) 2 ] 2 N N CH 3 N Au C 6 F 5 Cl C 6 F 5 N N CH 3 N Au C 6 F 5 C 6 F 5 C 6 F 5 CH 2 Cl 2, K 2 CO 3 [Au(C 6 F 5 ) 3 (tht)] 1 3 2 Au Cl C 6 F 5 C 6 F 5 Cl AuC 6 F 5 [AuCl(C 6 F 5 ) 2 ] 2 = C 6 F 5 H Scheme 1. Synthesis of gold(III) complexes 1 – 3 containing the acridine-based NHC ligand. tht = tetrahydrothiophene. This could be explained by the fact that the formation of the NHC ligand favored by the addition of the mild base K 2 CO 3 is able to displace the labile group (tht) to afford complexes 1 and 3 , or to promote the dimer cleavage to form complex 2 . However, very recently, Nolan and co-workers reported an interesting mechanistic pathway in which the Molecules 2022,27, 8289 4 of 16 aurate(I) anion and a mild base, such as K 2 CO 3 or NEt 3 , play a key role in the formation of the final [AuCl(NHC)] complex [ 31 ]. In this sense, and following the same hypothesis, a plausible reaction pathway is proposed for the formation of complexes 1 – 3 . Similarly, as observed for those gold(I) complexes previously reported [ 31 ], all of the intermediates ( Int1 – Int3 ) were detected by 1 H NMR spectroscopy in the crude reaction after just 10 min of the reaction (See Figures S4, S8 and S12). This is also associated with the higher stability of the intermediates compared to the corresponding precursors imidazolium chloride salts (Scheme 2) [31]. Molecules 2022, 27, x FOR PEER REVIEW 4 of 16 This could be explained by the fact that the formation of the NHC ligand favored by the addition of the mild base K2CO3 is able to displace the labile group (tht) to afford complexes 1 and 3, or to promote the dimer cleavage to form complex 2. However, very recently, Nolan and co-workers reported an interesting mechanistic pathway in which the aurate(I) anion and a mild base, such as K2CO3 or NEt3, play a key role in the formation of the final [AuCl(NHC)] complex [31]. In this sense, and following the same hypothesis, a plausible reaction pathway is proposed for the formation of complexes 1–3. Similarly, as observed for those gold(I) complexes previously reported [31], all of the intermediates (Int1–Int3) were detected by 1H NMR spectroscopy in the crude reaction after just 10 min of the reaction (See Figures S4, S8 and S12). This is also associated with the higher stability of the intermediates compared to the corresponding precursors imidazolium chloride salts (Scheme 2) [31]. Scheme 2. Synthesis of gold(III) complexes 1–3 containing the acridine-based NHC ligand. tht = tetrahydrothiophene. Because of the small size of the methyl group on the imidazolium salt, the approach of the aurate(III) anion could not be so sterically hindered to promote the formation of a concerted bond-making/bond-breaking situation leading to the corresponding Au(III)- NHC complexes (1, 2 or 3) and the spontaneous precipitation of KCl and KHCO3 (Scheme 2) [31]. The influence of the incorporation of the pentafluorophenyl group in the acridinebased NHC-gold complexes was initially evidenced by better solubility, even in less polar solvents. They displayed the following tendency in solubility: [AuCl3(NHC)] (1) < [AuCl(C6F5)2(NHC)] (2) < [Au(C6F5)3(NHC)] (3). For example, complex 1 only displayed moderate solubility in polar solvents such as DMSO or methanol, while complexes 2 and 3 presented excellent solubility properties in solvents such as acetone, acetonitrile, dichloromethane and even chloroform (75 mg/mL). As expected, all the complexes obtained shared some 1H NMR spectroscopy features due to the hydrogen signals assigned to the acridine-based NHC ligand. The absence of a singlet at 9.95 ppm corresponding to the NCHN imidazolium proton confirms the formation of the Au(III)-NHC complexes [9]. Although 13C-{1H} NMR spectroscopy is normally used to characterize NHC-metal complexes, the C2 carbenic carbon could not be identified, even when using two-dimensional NMR spectroscopy. This may be indicative of a rapid relaxation process of the carbon nuclei. In addition, the differences among their chemical shifts and their multiplicity K 2 CO 3 N N Me Au X X Y N Cl H N N Me N H Cl + N N Me Au X X Y N Cl H -KHCO 3 -KCl N N Me Au X X Y N C - O O OK + K + Intermediate X = Y = Cl (Int1) X = C 6 F 5 ; Y = Cl (Int2) X = Y = C 6 F 5 (Int3) [AuX 2 Y(tht)] or 1/ 2 [AuX 2 Y] 2 Complex X = Y = Cl (1) X = C 6 F 5 ; Y = Cl (2) X = Y = C 6 F 5 (3) Scheme 2. Mechanism for the synthesis of gold(III) complexes 1–3. Because of the small size of the methyl group on the imidazolium salt, the approach of the aurate(III) anion could not be so sterically hindered to promote the formation of a concerted bond-making/bond-breaking situation leading to the corresponding Au(III)-NHC complexes ( 1 , 2 or 3 ) and the spontaneous precipitation of KCl and KHCO 3 (Scheme 2) [ 31 ]. The influence of the incorporation of the pentafluorophenyl group in the acridinebased NHC-gold complexes was initially evidenced by better solubility, even in less polar solvents. They displayed the following tendency in solubility: [AuCl 3 (NHC)] ( 1 ) < [AuCl(C 6 F 5 ) 2 (NHC)] ( 2 ) < [Au(C 6 F 5 ) 3 (NHC)] ( 3 ). For example, complex 1 only displayed moderate solubility in polar solvents such as DMSO or methanol, while complexes 2 and 3 presented excellent solubility properties in solvents such as acetone, acetonitrile, dichloromethane and even chloroform (75 mg/mL). As expected, all the complexes obtained shared some 1 H NMR spectroscopy features due to the hydrogen signals assigned to the acridine-based NHC ligand. The absence of a singlet at 9.95 ppm corresponding to the NCHN imidazolium proton confirms the formation of the Au(III)-NHC complexes [ 9 ]. Although 13 C-{ 1 H} NMR spectroscopy is normally used to characterize NHC-metal complexes, the C2 carbenic carbon could not be identified, even when using two-dimensional NMR spectroscopy. This may be indicative of a rapid relaxation process of the carbon nuclei. In addition, the differences among their chemical shifts and their multiplicity depend on both the deuterated solvent used and the gold(III) center, which for the obtained complexes had a different extent of functionalization due to the presence of pentafluorophenyl groups [ 27 , 29 ]. Nevertheless, in the mass spectra, the molecular peak [M-2Cl + H]+ appeared at m/z= 492.0562, and [M-Cl] + appeared at m/z790.0523 for complexes 1 and 2 , Molecules 2022,27, 8289 5 of 16 respectively. For complex 3, the peak observed at m/z= 958.0623 was assigned to [M + H] + (See Figures S7, S11 and S15). It is noteworthy that the dimer cleavage promoted by the NHC ligand to form complex 2 only gave the cis isomer (Scheme 1). This could be associated with the cis conformation of the dimer precursor [Au( µ -Cl)(C 6 F 5 ) 2 ] 2 and also due to the steric hindrance imposed by the carbene ligand (Scheme 1). Consequently, both gold(III) complexes ( 2 and 3 ) displayed two sets of signals in the 19 F NMR spectra corresponding to the ortho-, metaand parafluorine atoms in the –C 6 F 5 group ranging from − 120.90 to − 164.70 ppm, similar to those reported in the literature [ 27 ]. According to the electronic spectra solved herein, all of the gold(III) complexes displayed structured bands with two maxima at 250 and 360 nm (see Figures S1–S3). These bands can be assigned to π→π *- or n →π *-type transitions within the acridine chromophore, which are in good agreement with the spectra previously obtained for the acridine-based NHC-gold(I) complexes and other reported data for acridine-based derivatives [ 32 ]. It is noteworthy that none of the complexes displayed photoluminescent properties. The high electrophilicity of the gold(III) ion could be associated with the luminescence quenching observed for these species, which can promote d–d transitions that are energetically close to emissive IL or MLCT states [ 3 ]. This fact makes these Au(III) derivatives inappropriate for use as biomarkers in living cell imaging agents. 2.2. X-ray Structure Analysis Single crystals of gold(III) complexes 1 and 2 were obtained and have been analyzed by X-ray diffraction. Their solid structures are depicted in Figures 3and 4. The crystal structure of complex 1 was monoclinic and contained two molecules per asymmetric unit. As expected for gold(III) complexes, the geometry around the metal center was square planar. On the other hand, the solid structure of complex 2 was triclinic and displays one molecule per asymmetric unit and was crystallized as an acetone adduct (Figure 4) with similar bond angles associated with a square planar geometry. Both complexes had the acridine-based NHC ligand in their structures where the carbenic carbon binds the Au(III) metal atom, in which the disposition of the acridine and imidazole rings was almost perpendicular (89.82 ◦ ) for one of the molecules in the asymmetric unit of complex 1 , whereas the torsion angle for complex 2was found to be 65.78◦(see Table 1). Molecules 2022, 27, x FOR PEER REVIEW 5 of 16 depend on both the deuterated solvent used and the gold(III) center, which for the obtained complexes had a different extent of functionalization due to the presence of pentafluorophenyl groups [27,29]. Nevertheless, in the mass spectra, the molecular peak [M2Cl + H] + appeared at m/z = 492.0562, and [M-Cl] + appeared at m/z 790.0523 for complexes 1 and 2, respectively. For complex 3, the peak observed at m/z = 958.0623 was assigned to [M + H] + (See Figures S7, S11 and S15). It is noteworthy that the dimer cleavage promoted by the NHC ligand to form complex 2 only gave the cis isomer (Scheme 1). This could be associated with the cis conformation of the dimer precursor [Au(μ-Cl)(C 6 F 5 ) 2 ] 2 and also due to the steric hindrance imposed by the carbene ligand (Scheme 1). Consequently, both gold(III) complexes (2 and 3) displayed two sets of signals in the 19 F NMR spectra corresponding to the ortho-, metaand para-fluorine atoms in the –C 6 F 5 group ranging from −120.90 to −164.70 ppm, similar to those reported in the literature.[27] According to the electronic spectra solved herein, all of the gold(III) complexes displayed structured bands with two maxima at 250 and 360 nm (see Figures S1–S3). These bands can be assigned to π→π*- or n→π*-type transitions within the acridine chromophore, which are in good agreement with the spectra previously obtained for the acridine-based NHC-gold(I) complexes and other reported data for acridine-based derivatives.[32] It is noteworthy that none of the complexes displayed photoluminescent properties. The high electrophilicity of the gold(III) ion could be associated with the luminescence quenching observed for these species, which can promote d–d transitions that are energetically close to emissive IL or MLCT states.[3] This fact makes these Au(III) derivatives inappropriate for use as biomarkers in living cell imaging agents. 2.2. X-ray Structure Analysis Single crystals of gold(III) complexes 1 and 2 were obtained and have been analyzed by X-ray diffraction. Their solid structures are depicted in Figures 3 and 4. The crystal structure of complex 1 was monoclinic and contained two molecules per asymmetric unit. As expected for gold(III) complexes, the geometry around the metal center was square planar. On the other hand, the solid structure of complex 2 was triclinic and displays one molecule per asymmetric unit and was crystallized as an acetone adduct (Figure 4) with similar bond angles associated with a square planar geometry. Both complexes had the acridine-based NHC ligand in their structures where the carbenic carbon binds the Au(III) metal atom, in which the disposition of the acridine and imidazole rings was almost perpendicular (89.82°) for one of the molecules in the asymmetric unit of complex 1, whereas the torsion angle for complex 2 was found to be 65.78° (see Table 1). Figure 3. Solid structure of complex 1 with 50% probability ellipsoids. Hydrogen atoms and dichloromethane molecules are omitted for clarity. Selected bond lengths (Å) and angles (°): Au(1)- C(1) 2.007(6), Cl(2)-Au(1)-Cl(1) 91.41(7), Cl(3)-Au(1)-Cl(1) 90.93(7), C(1)-Au(1)-Cl(2) 89.52(17), C(1)- Au(1)-Cl(3) 88.15(17). Figure 3. Solid structure of complex 1 with 50% probability ellipsoids. Hydrogen atoms and dichloromethane molecules are omitted for clarity. Selected bond lengths (Å) and angles ( ◦ ): Au(1)- C(1) 2.007(6), Cl(2)-Au(1)-Cl(1) 91.41(7), Cl(3)-Au(1)-Cl(1) 90.93(7), C(1)-Au(1)-Cl(2) 89.52(17), C(1)- Au(1)-Cl(3) 88.15(17). Molecules 2022,27, 8289 6 of 16 Molecules 2022, 27, x FOR PEER REVIEW 6 of 16 Figure 4. Solid structure of complex 2 with 50% probability ellipsoids. Hydrogen atoms and acetone molecule are omitted for clarity. Selected bond lengths (Å) and angles (°): Au(1)-C(1) 2.047(2) (C(1)- Au(1)-Cl(1) 89.18(6), C(18)-Au(1)-C(1) 88.27(8), C(18)-Au(1)-C(24) 91.06(9), C(24)-Au(1)-Cl(1) 91.51(6). Table 1. Selected Bond Lengths (Å), Angles (°), Torsion Angles (°), Interplanar Distance (Å) and Centroid–Centroid Distance for Complexes 1 and 2. Compound 1 2 Au–C carb 2.007 (6) 2.047 (2) Au–X trans 2.3198 (15) 2.053 (2) Au–Cl(1) 2.3198 (15) 2.3323 (6) C carb –Au–X trans 178.74 (17) 179.03 (8) torsion angle [a] 89.82 65.78 interplanar distance [b] 3.517 3.504 Centroid–centroid distance 3.583 3.701 aromatic contact py-py py-py [a] Torsion angle between imidazole and acridine rings. [b] Interplanar distance between acridine rings. Similar to other gold(I) complexes containing the acridine-based NHC ligand previously reported,[26] a significant difference was found in the carbene–gold bond distances Au(1)–C(1) of 2.007(6) and 2.047(2) Å for complex 1 and 2, respectively, after swapping a chloride ligand for a pentafluorophenyl group. This is in good agreement with the great trans-influence associated with the C 6 F 5 - fragment as compared to the chloride ligand. This is also seen when comparing Au1-Cl1 bond distances in complexes 1 and 2 (2.3198(15) vs. 2.3323(6)), in which the C 6 F 5 - anion seems to be a greater trans-influence group than even the acridine-based NHC ligand. As expected, the presence of the acridine moiety in both complexes promoted the formation of π···π stacking interactions between the acridine rings of different molecules [9,33]. Calculation of planes and centroids allowed the determination of a py-py contact between the two aromatic rings with interplanar and centroid–centroid distances of 3.517 and 3.583 and 3.504 and 3.701 Å, for complexes 1 and 2, respectively (Figures 5 and 6). The centroid–centroid distances found are in good agreement with those obtained for other metal-based analogs containing an acridine moiety ranging from 3.585 to 3.820 Å [9,34,35]. Parallel disposition is normally associated with a displacement angle between the acridine rings. In this case, complex 2 showed a displacement angle of 18.77°, while that of complex 1 was only 11.10°, the smallest value found for the acridine-based gold-NHC complex. Although complex 3 was identified by NMR spectroscopy and mass spectrometry, no single crystals suitable for X-ray diffraction analysis were obtained. Figure 4. Solid structure of complex 2 with 50% probability ellipsoids. Hydrogen atoms and acetone molecule are omitted for clarity. Selected bond lengths (Å) and angles ( ◦ ): Au(1)-C(1) 2.047(2) (C(1)-Au(1)-Cl(1) 89.18(6), C(18)-Au(1)-C(1) 88.27(8), C(18)-Au(1)-C(24) 91.06(9), C(24)-Au(1)-Cl(1) 91.51(6). Table 1. Selected Bond Lengths (Å), Angles ( ◦ ), Torsion Angles ( ◦ ), Interplanar Distance (Å) and Centroid–Centroid Distance for Complexes 1and 2. Compound 1 2 Au–Ccarb 2.007 (6) 2.047 (2) Au–Xtrans 2.3198 (15) 2.053 (2) Au–Cl(1) 2.3198 (15) 2.3323 (6) Ccarb–Au–Xtrans 178.74 (17) 179.03 (8) torsion angle [a] 89.82 65.78 interplanar distance [b] 3.517 3.504 Centroid–centroid distance 3.583 3.701 aromatic contact py-py py-py [a] Torsion angle between imidazole and acridine rings. [b] Interplanar distance between acridine rings. Similar to other gold(I) complexes containing the acridine-based NHC ligand previously reported [ 26 ], a significant difference was found in the carbene–gold bond distances Au(1)–C(1) of 2.007(6) and 2.047(2) Å for complex 1 and 2 , respectively, after swapping a chloride ligand for a pentafluorophenyl group. This is in good agreement with the great trans-influence associated with the C 6 F 5 - fragment as compared to the chloride ligand. This is also seen when comparing Au1-Cl1 bond distances in complexes 1 and 2 (2.3198(15) vs. 2.3323(6)), in which the C 6 F 5 - anion seems to be a greater trans-influence group than even the acridine-based NHC ligand. As expected, the presence of the acridine moiety in both complexes promoted the formation of π···π stacking interactions between the acridine rings of different molecules [9,33] . Calculation of planes and centroids allowed the determination of a py-py contact between the two aromatic rings with interplanar and centroid–centroid distances of 3.517 and 3.583 and 3.504 and 3.701 Å, for complexes 1 and 2 , respectively (Figures 5and 6). The centroid– centroid distances found are in good agreement with those obtained for other metal-based analogs containing an acridine moiety ranging from 3.585 to 3.820 Å [ 9 , 34 , 35 ]. Parallel disposition is normally associated with a displacement angle between the acridine rings. In this case, complex 2 showed a displacement angle of 18.77 ◦ , while that of complex 1 was only 11.10 ◦ , the smallest value found for the acridine-based gold-NHC complex. Although complex 3 was identified by NMR spectroscopy and mass spectrometry, no single crystals suitable for X-ray diffraction analysis were obtained. Molecules 2022,27, 8289 7 of 16 Molecules 2022, 27, x FOR PEER REVIEW 7 of 16 Figure 5. Offset π···π stacking interactions of complex 1. Figure 6. Offset π···π stacking interactions of complex 2. 2.3. Synthesis of the 1,3-Bis-(2,6-diisopropylphenyl)imidazolinium tetrakis(pentafluorophenyl)aurate(III) Salt To explore the reactivity of the precursor [Au(C6F5)3(tht)] against a different NHC ligand, we prepared a complex with the general formula [Au(C6F5)3(NHC)] using the bulkier ligand 1,3-Bis-(2,6-diisopropylphenyl)imidazolidine-2-ylidene (SIPr), which is one of the bulkiest ligands among the NHCs commonly used in different applications.[36] To obtain the desired complex, [Au(C6F5)3(SIPr)], and the acridine-based Au(III)-NHC complexes, a reaction between the imidazolinium salt [SIPr-H]Cl and [Au(C6F5)3(tht)] was induced in the presence of K2CO3. However, the presence of the bulky NHC ligand (SIPr) possibly prevented the formation of the desired gold-NHC complex due to the steric hindrance (Scheme 3). In fact, the direct reaction of the free SIPr with [Au(C6F5)3(tht)] did not afford the complex [Au(C6F5)3(SIPr)] 4*. The limitations in the formation of the organometallic complex (4*) were evidenced by the gradual increase in the purple color in the reaction mixture, which is characteristic of the formation of gold nanoparticles and the final formation of species 4. As reported by Nevado and co-workers, this could be rationalized as the result of three steps. First, the detection of decafluorobiphenyl (C6F5–C6F5) in the initial crude reaction suggests that a cross-coupling reaction can be achieved via reductive elimination of an increasingly small amount of the tris(pentafluorophenyl)gold(III) intermediate promoted by the presence of K2CO3 to afford [SIPr-H][Au(C6F5)Cl] [37]. Subsequently, the excess of K2CO3 remaining in the crude reaction could also lead to the formation of gold nanoparticles and scrambling of ligands (C6F5− and unknown side products) through a disproportion reaction of the chloridopentafluorophenylaurate(I) anion [37,38]. Finally, the unreacted intermediate ([SIPr-H][Au(C6F5)Cl]) remaining in the reaction medium is able to rapidly trap a C6F5− fragment to form the tetrakis(pentafluorophenyl)aurate(III) [Au(C6F5)4]− anion, which is stabilized by the imidazolinium cation to afford 4 (Scheme 3). Figure 5. Offset π···πstacking interactions of complex 1. Molecules 2022, 27, x FOR PEER REVIEW 7 of 16 Figure 5. Offset π···π stacking interactions of complex 1. Figure 6. Offset π···π stacking interactions of complex 2. 2.3. Synthesis of the 1,3-Bis-(2,6-diisopropylphenyl)imidazolinium tetrakis(pentafluorophenyl)aurate(III) Salt To explore the reactivity of the precursor [Au(C6F5)3(tht)] against a different NHC ligand, we prepared a complex with the general formula [Au(C6F5)3(NHC)] using the bulkier ligand 1,3-Bis-(2,6-diisopropylphenyl)imidazolidine-2-ylidene (SIPr), which is one of the bulkiest ligands among the NHCs commonly used in different applications.[36] To obtain the desired complex, [Au(C6F5)3(SIPr)], and the acridine-based Au(III)-NHC complexes, a reaction between the imidazolinium salt [SIPr-H]Cl and [Au(C6F5)3(tht)] was induced in the presence of K2CO3. However, the presence of the bulky NHC ligand (SIPr) possibly prevented the formation of the desired gold-NHC complex due to the steric hindrance (Scheme 3). In fact, the direct reaction of the free SIPr with [Au(C6F5)3(tht)] did not afford the complex [Au(C6F5)3(SIPr)] 4*. The limitations in the formation of the organometallic complex (4*) were evidenced by the gradual increase in the purple color in the reaction mixture, which is characteristic of the formation of gold nanoparticles and the final formation of species 4. As reported by Nevado and co-workers, this could be rationalized as the result of three steps. First, the detection of decafluorobiphenyl (C6F5–C6F5) in the initial crude reaction suggests that a cross-coupling reaction can be achieved via reductive elimination of an increasingly small amount of the tris(pentafluorophenyl)gold(III) intermediate promoted by the presence of K2CO3 to afford [SIPr-H][Au(C6F5)Cl] [37]. Subsequently, the excess of K2CO3 remaining in the crude reaction could also lead to the formation of gold nanoparticles and scrambling of ligands (C6F5− and unknown side products) through a disproportion reaction of the chloridopentafluorophenylaurate(I) anion [37,38]. Finally, the unreacted intermediate ([SIPr-H][Au(C6F5)Cl]) remaining in the reaction medium is able to rapidly trap a C6F5− fragment to form the tetrakis(pentafluorophenyl)aurate(III) [Au(C6F5)4]− anion, which is stabilized by the imidazolinium cation to afford 4 (Scheme 3). Figure 6. Offset π···πstacking interactions of complex 2. 2.3. Synthesis of the 1,3-Bis-(2,6-diisopropylphenyl)imidazolinium tetrakis(pentafluorophenyl)aurate(III) Salt To explore the reactivity of the precursor [Au(C 6 F 5 ) 3 (tht)] against a different NHC ligand, we prepared a complex with the general formula [Au(C 6 F 5 ) 3 (NHC)] using the bulkier ligand 1,3-Bis-(2,6-diisopropylphenyl)imidazolidine-2-ylidene (SIPr), which is one of the bulkiest ligands among the NHCs commonly used in different applications [ 36 ]. To obtain the desired complex, [Au(C 6 F 5 ) 3 (SIPr)], and the acridine-based Au(III)-NHC complexes, a reaction between the imidazolinium salt [SIPr-H]Cl and [Au(C 6 F 5 ) 3 (tht)] was induced in the presence of K 2 CO 3 . However, the presence of the bulky NHC ligand (SIPr) possibly prevented the formation of the desired gold-NHC complex due to the steric hindrance (Scheme 3). In fact, the direct reaction of the free SIPr with [Au(C 6 F 5 ) 3 (tht)] did not afford the complex [Au(C 6 F 5 ) 3 (SIPr)] 4* . The limitations in the formation of the organometallic complex ( 4* ) were evidenced by the gradual increase in the purple color in the reaction mixture, which is characteristic of the formation of gold nanoparticles and the final formation of species 4 . As reported by Nevado and co-workers, this could be rationalized as the result of three steps. First, the detection of decafluorobiphenyl (C 6 F 5 –C 6 F 5 ) in the initial crude reaction suggests that a cross-coupling reaction can be achieved via reductive elimination of an increasingly small amount of the tris(pentafluorophenyl)gold(III) intermediate promoted by the presence of K 2 CO 3 to afford [SIPr-H][Au(C 6 F 5 )Cl] [ 37 ]. Subsequently, the excess of K 2 CO 3 remaining in the crude reaction could also lead to the formation of gold nanoparticles and scrambling of ligands (C 6 F 5− and unknown side products) through a disproportion reaction of the chloridopentafluorophenylaurate(I) anion [ 37 , 38 ]. Finally, the unreacted intermediate ([SIPr-H][Au(C 6 F 5 )Cl]) remaining in the reaction medium is able to rapidly trap a C 6 F 5− fragment to form the tetrakis(pentafluorophenyl)aurate(III) [Au(C 6 F 5 ) 4 ] − anion, which is stabilized by the imidazolinium cation to afford 4 (Scheme 3). Molecules 2022,27, 8289 8 of 16 Molecules 2022, 27, x FOR PEER REVIEW 8 of 16 Scheme 3. Proposed reaction pathway for the formation of compound 4. The 1 H NMR spectrum of gold–imidazolinium salt 4 showed a singlet at 7.72 ppm characteristic of the N(N)C-H proton of the N-heterocycle, which was strongly upfielded with respect to that observed for the precursor [SIPr-H]Cl (9.55 ppm) and also the intermediate, which appeared at 8.05 ppm (see Figure S13). The other signals did not show significant changes relative to the chemical shifts in the 1 H NMR spectra. The [Au(C 6 F 5 ) 4 ] − counteranion showed two multiplets at −122.36 and −162.53 ppm and a triplet at −159.50 ppm in the 19 F NMR spectrum with 3 J F-F = 19.8 Hz. Single crystals of gold–imidazolinium salt 4 suitable for X-ray diffraction analysis were obtained, the solid structure of which is depicted in Figure 7. Although there are a few examples of salts (including imidazolium) being stabilized by gold(III) anions[39–44], to the best of our knowledge, this is the first solid-state structure of an imidazolinium cation stabilized by an aurate(III) anion reported in the literature. Figure 7. Solid structure of gold–imidazolium salt 4 with 50% probability ellipsoids. Selected bond lengths (Å) and angles (°): N(1)-C(1) 1.308(4), N(2)-C(1) 1.297(4), Au(1)-C(28) 2.044(3), Au(1)-C(34) 2.053(3), Au(1)-C40 2.059(3), Au(1)-C(46) 2.061(3); C(28)-Au(1)-C(34) 176.43(12), C(28)-Au(1)-C(40) 88.06(12), C(28)-Au(1)-C(46) 91.90(12), C(34)-Au(1)-C(40) 88.84(12), C(34)-Au(1)-C(46) 91.33(12), C(40)-Au(1)-C(46) 175.47(12). As observed, the bond angles were close to those found for gold(III) centers with square planar geometry. Compound 4 crystallized in the monoclinic space group P2(1)/c, with Au-C bond lengths and angles similar to those found for other previously reported Scheme 3. Proposed reaction pathway for the formation of compound 4 . Complex 4* represents the expected compound. The 1 H NMR spectrum of gold–imidazolinium salt 4 showed a singlet at 7.72 ppm characteristic of the N(N)C-H proton of the N-heterocycle, which was strongly upfielded with respect to that observed for the precursor [SIPr-H]Cl (9.55 ppm) and also the intermediate, which appeared at 8.05 ppm (see Figure S13). The other signals did not show significant changes relative to the chemical shifts in the 1 H NMR spectra. The [Au(C 6 F 5 ) 4 ] − counteranion showed two multiplets at − 122.36 and − 162.53 ppm and a triplet at − 159.50 ppm in the 19F NMR spectrum with 3JF-F = 19.8 Hz. Single crystals of gold–imidazolinium salt 4 suitable for X-ray diffraction analysis were obtained, the solid structure of which is depicted in Figure 7. Although there are a few examples of salts (including imidazolium) being stabilized by gold(III) anions [ 39 – 44 ], to the best of our knowledge, this is the first solid-state structure of an imidazolinium cation stabilized by an aurate(III) anion reported in the literature. Molecules 2022, 27, x FOR PEER REVIEW 8 of 16 Scheme 3. Proposed reaction pathway for the formation of compound 4. The 1 H NMR spectrum of gold–imidazolinium salt 4 showed a singlet at 7.72 ppm characteristic of the N(N)C-H proton of the N-heterocycle, which was strongly upfielded with respect to that observed for the precursor [SIPr-H]Cl (9.55 ppm) and also the intermediate, which appeared at 8.05 ppm (see Figure S13). The other signals did not show significant changes relative to the chemical shifts in the 1 H NMR spectra. The [Au(C 6 F 5 ) 4 ] − counteranion showed two multiplets at −122.36 and −162.53 ppm and a triplet at −159.50 ppm in the 19 F NMR spectrum with 3 J F-F = 19.8 Hz. Single crystals of gold–imidazolinium salt 4 suitable for X-ray diffraction analysis were obtained, the solid structure of which is depicted in Figure 7. Although there are a few examples of salts (including imidazolium) being stabilized by gold(III) anions[39–44], to the best of our knowledge, this is the first solid-state structure of an imidazolinium cation stabilized by an aurate(III) anion reported in the literature. Figure 7. Solid structure of gold–imidazolium salt 4 with 50% probability ellipsoids. Selected bond lengths (Å) and angles (°): N(1)-C(1) 1.308(4), N(2)-C(1) 1.297(4), Au(1)-C(28) 2.044(3), Au(1)-C(34) 2.053(3), Au(1)-C40 2.059(3), Au(1)-C(46) 2.061(3); C(28)-Au(1)-C(34) 176.43(12), C(28)-Au(1)-C(40) 88.06(12), C(28)-Au(1)-C(46) 91.90(12), C(34)-Au(1)-C(40) 88.84(12), C(34)-Au(1)-C(46) 91.33(12), C(40)-Au(1)-C(46) 175.47(12). As observed, the bond angles were close to those found for gold(III) centers with square planar geometry. Compound 4 crystallized in the monoclinic space group P2(1)/c, with Au-C bond lengths and angles similar to those found for other previously reported Figure 7. Solid structure of gold–imidazolium salt 4 with 50% probability ellipsoids. Selected bond lengths (Å) and angles ( ◦ ): N(1)-C(1) 1.308(4), N(2)-C(1) 1.297(4), Au(1)-C(28) 2.044(3), Au(1)-C(34) 2.053(3), Au(1)-C40 2.059(3), Au(1)-C(46) 2.061(3); C(28)-Au(1)-C(34) 176.43(12), C(28)-Au(1)-C(40) 88.06(12), C(28)-Au(1)-C(46) 91.90(12), C(34)-Au(1)-C(40) 88.84(12), C(34)-Au(1)-C(46) 91.33(12), C(40)- Au(1)-C(46) 175.47(12). As observed, the bond angles were close to those found for gold(III) centers with square planar geometry. Compound 4 crystallized in the monoclinic space group P2(1)/c, Molecules 2022,27, 8289 9 of 16 with Au-C bond lengths and angles similar to those found for other previously reported tetrakis(pentafluorophenyl)gold(III) derivatives [ 39 ]. In the crystal packing (Figure 8), the imidazolinium cations were arranged in coupled rows along the b axis, as if related by a center of symmetry, with the anionic [Au(C 6 F 5 ) 4 ] − units filling the gaps between the rows. Molecules 2022, 27, x FOR PEER REVIEW 9 of 16 tetrakis(pentafluorophenyl)gold(III) derivatives.[39] In the crystal packing (Figure 8), the imidazolinium cations were arranged in coupled rows along the b axis, as if related by a center of symmetry, with the anionic [Au(C6F5)4]− units filling the gaps between the rows. Figure 8. View along b axis of the crystal packing of 4. Several attempts to use the [IPr-H]Cl salt under the same reaction conditions were unsuccessful due to the presence of a mixture of products that were very difficult to separate by means of standard separation techniques. 2.4. Computational Studies To understand the reactivity observed, calculations were carried out based on the framework of the density functional theory (DFT).[45] Geometry optimization calculations were performed on the reactants, products and intermediates associated with the formation of complexes 3 and 4, allowing us to obtain more stable structures and the corresponding reaction energies (see SI for computational details). The intermediates’ and products’ relative electronic energies were estimated as ∆Eint/prod = Eintermediate/product − ENHC − EAu-complex. As shown in Table 2, all energies indicated a thermodynamically favorable intermediate (Int4) and product formation ∆Eint < −47 kcal·mol−1 and ∆Eprod < −90 kcal·mol−1, respectively. It is worth mentioning that thermal and entropic corrections of the electronic energies did not have a significant impact on the relative energies (Tables S7 and S8). The amplitudes of these formation energies are characteristic of comparable gold-NHC complexes reported in the literature [31,46]. Because the reaction for the formation of complex 4 differed from that occurring for the formation of complex 3, we also considered the formation of a hypothetical complex [Au(C6F5)3(SIPr)] (4*), Scheme 3. The lower ∆Eprod of 4* (−114.6 kcal·mol−1) suggests its formation over that of the aurate(III) anion in salt 4. However, the alternative formation of Int5 after reductive elimination of the [Au(C6F5)3Cl]− complex at −69.2 kcal·mol−1, as well as its lesser steric hindrance, likely leads to a lower barrier to reaching 4. Hence, we expect the detection of 4 as the main product of the reaction mixture, as it was observed in our experiments. Figure 8. View along b axis of the crystal packing of 4. The axis are represented as a, b, c. Several attempts to use the [IPr-H]Cl salt under the same reaction conditions were unsuccessful due to the presence of a mixture of products that were very difficult to separate by means of standard separation techniques. 2.4. Computational Studies To understand the reactivity observed, calculations were carried out based on the framework of the density functional theory (DFT) [ 45 ]. Geometry optimization calculations were performed on the reactants, products and intermediates associated with the formation of complexes 3 and 4 , allowing us to obtain more stable structures and the corresponding reaction energies (see SI for computational details). The intermediates’ and products’ relative electronic energies were estimated as ∆ E int/prod =E intermediate/product − E NHC − E Au-complex . As shown in Table 2, all energies indicated a thermodynamically favorable intermediate ( Int4 ) and product formation ∆ E int < − 47 kcal · mol −1 and ∆Eprod <−90 kcal·mol−1 , respectively. It is worth mentioning that thermal and entropic corrections of the electronic energies did not have a significant impact on the relative energies (Tables S7 and S8). The amplitudes of these formation energies are characteristic of comparable gold-NHC complexes reported in the literature [ 31 , 46 ]. Because the reaction for the formation of complex 4 differed from that occurring for the formation of complex 3 , we also considered the formation of a hypothetical complex [Au(C 6 F 5 ) 3 (SIPr)] ( 4* ), Scheme 3. The lower ∆ E prod of 4* ( − 114.6 kcal · mol −1 ) suggests its formation over that of the aurate(III) anion in salt 4 . However, the alternative formation of Int5 after reductive elimination of the [Au(C 6 F 5 ) 3 Cl] − complex at − 69.2 kcal · mol −1 , as well as its lesser steric hindrance, likely leads to a lower barrier to reaching 4 . Hence, we expect the detection of 4 as the main product of the reaction mixture, as it was observed in our experiments. Molecules 2022,27, 8289 16 of 16 63. Barone, V.; Cossi, M. Conductor Solvent Model. J. Phys. Chem. A 1998,102, 1995–2001. [CrossRef] 64. Hanwell, M.D.; Curtis, D.E.; Lonie, D.C.; Vandermeerschd, T.; Zurek, E.; Hutchison, G.R. Avogadro: An Advanced Semantic Chemical Editor, Visualization, and Analysis Platform. J. Cheminform. 2012,4, 1–17. [CrossRef] [PubMed]